GAIL prep: Weekend batch (Sat & Sun 10 AM-12 PM IST)

What you'll learn

1. 1. Week 1: Orientation + Module 1: Gen AI, Beyond the ChatbotHow the exam works: format, 4 domains and weightings, and the rebranded product names (Agent Platform, Agent Studio, Agent Search, Gemini Notebook) · The four ways to use gen AI: create, summarize, discover, automate · Gen AI vs traditional AI; where gen AI creates business value · Gemini app basics; choosing work tasks for gen AI
2. 2. Week 2: Module 2: Unlock Foundational Concepts (I): AI, ML and dataAI vs ML vs deep learning vs gen AI; NLP · Supervised, unsupervised and reinforcement learning · Structured vs unstructured data, labelled vs unlabelled data, data quality and accessibility · The ML lifecycle (ingest, prepare, train, deploy, manage) and the Google Cloud tools for each stage
3. 3. Week 3: Module 2: Unlock Foundational Concepts (II): foundation models and their limitsFoundation models, LLMs, multimodal and diffusion models, prompt tuning, fine-tuning · Choosing a foundation model: modality, context window, security, availability and reliability, cost, performance, customization · Limits: data dependency, knowledge cutoff, bias, fairness, hallucinations, edge cases · Fixes: grounding, RAG, prompt engineering, fine-tuning, human in the loop (official 'Summary of techniques' handout)
4. 4. Week 4: Module 3: Navigate the LandscapeThe 5 layers: infrastructure, models, platforms, agents, applications · Google models: Gemini, Gemma, Imagen / Nano Banana, Veo; first-party, open and third-party models in Model Garden · AI-optimized infrastructure: TPUs, GPUs, AI Hypercomputer; Google's open approach and enterprise readiness · AI at the edge: LiteRT, Gemini Nano
5. 5. Week 5: Module 4: Gen AI Apps, Transform Your Work (I): productivityGemini app and Gemini Advanced (Google AI plans); Gems · Google Workspace with Gemini: Gmail, Docs, Sheets, Slides, Meet, Vids · Gemini Enterprise: enterprise search, Gemini Notebook API, custom agents · Gemini Notebook / NotebookLM Enterprise for document understanding
6. 6. Week 6: Module 4 (II): Prompt engineering and sampling parametersZero-, one- and few-shot, role prompting, prompt chaining · Chain-of-thought, ReAct, metaprompting; reusing prompts, saved info · Sampling parameters: token count, temperature, top-p, safety settings, output length · System instructions
7. 7. Week 7: Module 4 (III): Semantic search, vector databases, grounding and RAGSemantic search and embeddings; vector databases · Grounding with first-party enterprise data, third-party data and world data · RAG steps: retrieve, augment, generate, iterate · Google grounding offerings: pre-built RAG with Agent Search, RAG APIs, Grounding with Google Search · Transparency and trust (Google Cloud transparency resources)
8. 8. Week 8: Customer experience and pre-built AI APIs + MOCK EXAM 1Customer Engagement Suite: Conversational Agents, Agent Assist, Conversational Insights, Contact Center as a Service · External search: Agent Search, Google Search · Pre-built APIs: Speech-to-Text, Text-to-Speech, Translation, Document Translation, Document AI, Vision, Video Intelligence, Natural Language; API Library
9. 9. Week 9: Module 5: Gen AI Agents, Transform your Organization (I): agents and toolingWhat an agent is; parts: model, tools, reasoning loop · Deterministic, generative and hybrid agents · Agent tooling: extensions, functions, data stores, plugins · Agent Studio vs Google AI Studio: when to use which (official comparison handout)
10. 10. Week 10: Module 5 (II): Agent Platform for builders and running models in productionGemini Enterprise Agent Platform (formerly Vertex AI): Model Garden, Model Builder, AutoML, Agent Search, custom agents · MLOps: Model Registry, versioning, performance tracking, drift monitoring (Model Monitoring), Feature Store, Pipelines, automatic model upgrades, KPIs · Building apps: Gemini API keys, Cloud Run / Cloud Run functions, Apps Script, AppSheet · Agent Studio quickstart: send text prompts to Gemini
11. 11. Week 11: Business strategy, secure AI and responsible AI + MOCK EXAM 2Choosing a gen AI solution: needs (scale, customization, interaction, privacy, latency, connectivity) and resources (people, money, time) · Top-down + bottom-up strategy, change management, measuring impact and ROI · Secure AI: SAIF, secure-by-design infrastructure, IAM, Security Command Center, workload monitoring · Responsible AI: transparency, privacy (anonymization, pseudonymization), bias, fairness, accountability, explainability
12. 12. Week 12: Final revision, exam registration and CAPSTONE DEMO DAYRevision driven by each learner's mock-exam weak areas · Official sample questions walkthrough · Registration step by step: certification page, then Register; choosing online-proctored (check the online testing requirements) or a Pearson VUE test centre; ID and room rules · Exam-day strategy for 50-60 questions in 90 minutes

Final project demo: Gen AI Business Proposal + Working Agent/RAG Demo. Pick a real problem from your organization or industry (e.g. HR policy assistant, IT helpdesk triage, insurance claim document Q&A). Write a 6-8 page proposal covering the use case, value hypothesis and KPIs, the solution layer (Workspace with Gemini / Gemini Enterprise / custom agent on Agent Platform) and why, model choice with trade-offs, grounding/RAG design, security (SAIF, IAM), responsible AI and privacy, rollout and change plan, and cost/ROI. Build a working prototype in Google AI Studio, Agent Studio, Gemini Notebook or Gemini Enterprise: an agent grounded on sample documents that cites its sources and refuses out-of-scope questions.